Related Experiment Video
Updated: May 13, 2026
Identifying Frailty Using Point-of-Care Ultrasonography: Image Acquisition and Assessment
Published on: July 26, 2024
Superior predictive performance of the Risk Analysis Index over the modified Frailty Index-5 in patients undergoing
Cameron J Sabet1, Bhav Jain2, Bara M Hammadeh3
1Georgetown University Medical Center, Washington D.C., USA.
Background:
Anterior cervical procedures in octogenarians carry substantial perioperative risks, underscoring the need for accurate preoperative assessment. Frailty, defined as diminished physiological reserve, is a recognized predictor of adverse surgical outcomes. Two validated measures-the modified 5-Factor Frailty Index (mFI-5) and the Risk Analysis Index (RAI)-are widely used, but their comparative predictive performance in this population is not well established. We aim to compare the predictive performance of mFI-5 versus RAI for 30-day postoperative outcomes following elective anterior cervical procedures in octogenarians.
Methods:
We conducted a retrospective cohort study using ACS-NSQIP data (2015-2021) of patients aged 80-89 years undergoing elective anterior cervical procedures. The primary outcome was 30-day mortality, with secondary outcomes of major complications, readmission, reoperation, and non-home discharge. Frailty was measured by mFI-5 and RAI, univariate AUROC comparisons were performed for each frailty index separately, followed by multivariable adjustment. Internal validation was performed using bootstrap resampling, with enhanced sensitivity analyses for mortality.
Results:
Among 870 patients, RAI classified 19.5% as frail versus 14.5% by mFI-5. In univariate analysis, RAI demonstrated superior discrimination for 30-day mortality (AUROC 0.830 vs. 0.514, p < 0.001) with an unadjusted OR of 1.31 (95%CI:1.14-1.50). However, given the low number of mortality events (n = 7), these findings should be interpreted with caution. RAI also showed better discrimination for non-home discharge (0.674 vs. 0.528, p < 0.001), readmission (0.607 vs. 0.504, p = 0.022), and extended length-of-stay (0.627 vs. 0.540, p = 0.003). After multivariable adjustment, RAI maintained superior predictive performance across outcomes. Bootstrap-validated AUROCs confirmed RAI superiority.
Conclusions:
RAI demonstrated superior discrimination compared to mFI-5 for predicting mortality and postoperative outcomes in octogenarians undergoing anterior cervical surgery. While RAI showed promising discrimination for mortality, these results require validation in larger cohorts due to the low event rate. Therefore, the RAI should be preferred over the mFI-5 for preoperative risk stratification in octogenarians undergoing anterior cervical surgery.
